Publisher’s version / Version de l'éditeur:
Chemical Communications, 47, 14, p. 4294–4296, 2011-04-14
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.
For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.
https://doi.org/10.1039/C0CC05501H
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Effect of alloying magnesium with chromium and vanadium on
hydrogenation kinetics studied with neutron reflectometry
Kalisvaart, Peter; Luber, Erik; Fritzsche, Helmut; Mitlin, David
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
NRC Publications Record / Notice d'Archives des publications de CNRC:
https://nrc-publications.canada.ca/eng/view/object/?id=11a4a235-47a8-416b-a9c8-c1bd0f284c1f
https://publications-cnrc.canada.ca/fra/voir/objet/?id=11a4a235-47a8-416b-a9c8-c1bd0f284c1f
4294 Chem. Commun.,2011, 47, 4294–4296 This journal is c The Royal Society of Chemistry 2011
Cite this:
Chem. Commun
., 2011, 47, 4294–4296
Effect of alloying magnesium with chromium and vanadium on
hydrogenation kinetics studied with neutron reflectometryw
Peter Kalisvaart,*
aErik Luber,
aHelmut Fritzsche
band David Mitlin
acReceived 10th December 2010, Accepted 10th February 2011 DOI: 10.1039/c0cc05501h
The effect of chromium and vanadium alloying on the hydrogenation of a magnesium thin film is studied by neutron reflectometry. Immediate formation of a blocking MgD2layer is observed in pure
Mg, however in the alloyed film deuteration is rapid and almost completely homogeneous.
Magnesium hydride has very high reversible storage capacity of 7.6 wt.% but suffers from extremely slow sorption kinetics and high stability (77 kJ mol1 H2). The kinetic limitations
are mostly due to extremely poor hydrogen diffusivity (B1020m2s1at 200 1C)1through the hydride. Formation of
a continuous layer of MgH2has often been observed to effectively
block further hydrogenation when the particles are too large.2–4 Although the stability is a very important issue, the vast majority of efforts are directed at improving the kinetics. One possibility is to try and eliminate the diffusion limitations by reducing the grain size and/or particle size, typically to 10 nm or less.5–7Another possibility is preparation of composite materials. If the additives preferably alloy with each other and not with Mg, they form a separate catalyst phase which acts as a diffusion pathway through hydrided material. Al–Ti8,9 and Fe–Ti9 have
already been proved to work very well.
The combination of Cr and V is expected to work in a similar manner as Fe–Ti as neither forms alloys with Mg, V has appreciable hydrogen solubility and diffusion is usually very rapid in bcc metals.10
Fig. 1 shows the absorption kinetics of the 1st, 5th, 10th and 20th cycle for a free-standing 1.5 mm thick cosputtered Mg–10%Cr–10%V film (ESIw). For this catalytic mixture, extremely fast absorption kinetics are observed. After 10 cycles, 90% of the total capacity is absorbed within 10 seconds, which is similar to what was observed for other Mg–TM cosputtered films.9
Hydrogen sorption measurements cannot distinguish the individual contributions of surface dissociation, diffusion and phase transformation to the ‘overall’ kinetics. Therefore, an
in situ technique with high spatial resolution is required to separate bulk (diffusion, nucleation) and surface (dissociation) effects in real-time. To that end, we used Neutron Reflectometry (NR) which allowed us to determine the hydrogen/deuterium concentration as a function of depth with nanometre resolution. We applied NR to study the deuterium absorption at room temperature of 5 nm Pd/5 nm CrV/50 nm Mg(–10%Cr–10%V)/10 nm Ta/Si, deposited on Si(100) substrates. Although interfacial effects between Mg and Pd may improve the desorption kinetics,11using a bilayer catalyst has been found by several authors to improve the absorption kinetics compared to having Pd in direct contact with a Mg(alloy) layer.12–15 Hence, our choice is for Pd/CrV in this study. Since deuterium can only enter the sample through the exposed bilayer catalyst, determination of the deuterium concen-tration depth-profile over time should elucidate the effects of alloying the Mg on the rate of deuterium penetration through the stack.
Fig. 2 shows the measured neutron reflectivity curves during absorption of the pure Mg (top panel) and Mg–10%Cr–10%V layers (bottom panel). A sensitive indicator of deuterium absorption is the critical scattering vector qc = 4(pNb)1/2
below which there is total reflection,12where N is the total number density of scatterers in the layer and b is the average scattering length. The product Nb is called the Scattering Length Density (SLD). Due to the large scattering length of deuterium (bD = 6.67 fm),16 the increase in qc is the first
Fig. 1 Absorption kinetics of a 1.5 mm Mg–10%Cr–10%V film for the 1st, 5th, 10th and 20th cycle at 200 1C. Absorption pressure was 2.5 bar.
a
University of Alberta, T6G 2V4, Edmonton, AB, Canada. E-mail: pkalisvaart@gmail.com
bNRC SIMS, Canadian Neutron Beam Centre, Chalk River,
ON K0J 1J0, Canada
cNational Institute of Nanotechnology, Edmonton, AB, Canada
wElectronic supplementary information (ESI) available: Experimental conditions, model parameters corresponding to all SLD profiles in Fig. 3, reflectivity curves that were not shown in Fig. 2 and AFM micrographs. See DOI: 10.1039/c0cc05501h
ChemComm
Dynamic Article Links
www.rsc.org/chemcomm
COMMUNICATION
Downloaded by CISTI Archive Access on 23 March 2011
Published on 02 March 2011 on http://pubs.rsc.org | doi:10.1039/C0CC05501H
This journal is c The Royal Society of Chemistry 2011 Chem. Commun., 2011, 47, 4294–4296 4295
indication that the layers are absorbing deuterium. This is clearly observed in Fig. 2 where qc increases monotonically
with time, both for the Mg and Mg–10%Cr–10%V layers. From the measured reflectivity data, SLD profiles were generated using Parratt’s recursive algorithm, which yields the SLD as a function of depth as shown in Fig. 3. Interfacial roughness is accounted for in this approach.17The solid lines in Fig. 2 are the reflectivity curves corresponding to these SLD profiles. As can be seen from Fig. 2, the simulated and experimental reflectivity curves agree quite well.
The SLD profiles as a function of time for the pure Mg layer are shown in the top panel of Fig. 3. In the top part of the Mg layer the SLD sharply increases immediately after absorption starts. After this, the rate of deuteration decreases significantly and even after 20 hours the layer is not completely absorbed. A clear gradient in the deuterium concentration, which is modeled as two separate layers with high interfacial roughness, is clearly present in the middle of the layer. This strongly indicates the formation of a blocking layer of MgD2.
The deuterium content in a layer can be calculated as:12
CD¼ SMþD Sp¼0 tMþD tM 1 b M bD ð1Þ
where CDis the deuterium concentration in D/Metal, SM+D
and Sp=0 are the SLD in loaded and as-deposited states,
respectively, and tM+D and tM are the corresponding
thicknesses. bMand bDare the scattering lengths of the metal
and deuterium, respectively. The D/M ratio in the Mg layer is estimated as 0.44, 1 and 1.2 after 1, 10 and 20 h, respectively. After 1 h, the D/M ratio in the Ta layer is only 0.2 and slowly increases to 0.5 after 20 h.
The interfacial roughness between the layers increases rapidly to a point at which the Pd and CrV layers are no longer distinguishable from one another. Therefore, a single layer with an SLD between that of Pd (4.02 106 A˚2) and CrV
(1.4 106A˚2) was used to fit the reflectivity curves, as can be seen from Fig. 3. The root-mean-square roughness (Rq) found
from the fit for the ‘catalyst’/D2and Mg/‘catalyst’ interfaces are
25 A˚ after 1 h and 32 and 34 A˚, respectively, after 20 h. Atomic force microscopy (AFM) was performed on the Mg layer, both in as-deposited and hydrogenated states to get a measured value for Rqof the catalyst/D2 interface. The Rq
values found for the as-deposited state are 9.1 and 9.2 A˚ from fitting the reflectivity curve and AFM measurement, respectively. After hydrogenation, the roughness measured with AFM had increased to 21.4 A˚ (ESIw). This is lower than the 32 A˚ found from the fit to the neutron data, but because of tip dilation effects in AFM, the value found with this technique will always be lower than the actual value.
Significant differences between the behavior of Mg–10%Cr–10%V and the pure Mg layer are immediately obvious from Fig. 3 (bottom panel). The Ta layer on the bottom of the stack attains its final SLD immediately after the first measurement at 10 minutes. Another remarkable feature in the SLD profiles is that the thickness of the Mg alloy layer is identical to the as-deposited state, despite the fact that the deuterium concentration increases to 0.25 and 0.5 D/M after 10 and 20 minutes, at 10 mbar pressure. The top part of the Mg–Cr–V layer has already reached its final SLD after only 30 minutes. There is a small gradient in the deuterium concen-tration, but there is no evidence of a blocking layer.
Fig. 2 Measured reflectivity curves (open symbols) and fits (solid lines) for a 50 nm pure Mg (top) and Mg–10%Cr–10%V (bottom) film in as-deposited and partially absorbed states at room temperature. PD2 for Mg was 50 mbar (ESIw), for Mg–10%Cr–10%V 10 mbar.
Fig. 3 SLD depth profiles for a pure Mg film in as-deposited state and as a function of time at 50 mbar absorption pressure (top panel) at room temperature. Same for Mg–10%Cr–10%V in the as-deposited state, after 10, 20, 30 min and 6.5 h at 10 mbar and finally 1 bar (bottom panel).
Downloaded by CISTI Archive Access on 23 March 2011
Published on 02 March 2011 on http://pubs.rsc.org | doi:10.1039/C0CC05501H
4296 Chem. Commun.,2011, 47, 4294–4296 This journal is c The Royal Society of Chemistry 2011
Between 20 and 30 minutes, the thickness of the Mg alloy layer increases from 554 A˚ to 710 A˚ and finally 718 A˚ when the layer is fully saturated, which is an expansion of 29.6%. The final D/M ratio in the Mg alloy layer calculated with eqn (1) is 1.54, which is close to the expected maximum value of 1.6 (=2 D/Mg). For the Pd and CrV layers, these values are 0.07 and 0.26, respectively, at 1 bar pressure. Thus, based on our measurements, a maximum 0.052 of the 1.54 D/M can be contributed by Cr and V.
Both films were investigated by X-ray diffraction in their as-deposited and hydrogenated states. As shown in Fig. 4, the layer is strongly (002) textured. The (101) peak was also detected by tilting the sample, enabling calculation of both lattice parameters as a = 3.17 and c = 5.12 A˚. The unit cell volume is B6% larger than expected based on a linear combination of the molar volumes of Mg, V and Cr. The Mg grain size is determined as 246 A˚ from the integral breadth of the (002) peak using the Scherrer equation (ESIw).
XRD measurements confirmed the formation of crystalline MgH2for the Mg layer which explains the blocking effect that
was observed in Fig. 3. For Mg–10%Cr–10%V, the position of the MgH2(110) reflection is identical, which means all Cr
and V segregate from Mg upon hydrogenation. The crystallite sizes of MgH2are 119 A˚ for pure Mg vs. 56 A˚ for the alloy
layer. The smaller grain size also makes the alloy film much smoother; Rqdetermined with AFM was only 6.1 A˚ (ESIw).
The Cr–V phase appears to limit the MgH2 grain size by
forming a networked structure along the grain boundaries. Such a network would serve as a rapid diffusion pathway for deuterium as well as providing a high number of heterogeneous nucleation sites for MgH2. This explains the absence of a blocking layer in
Mg–10%Cr–10%V and the almost homogeneous increase in SLD observed in Fig. 3. For the Mg layer, the only heterogeneous nucleation site is at the interface between the CrV and Mg layers leading to a blocking deuteride layer. Nucleation also appears more difficult there as the required pressure was 5 higher compared to Mg–10%Cr–10%V (ESIw).
Based on the XRD results, up to 6% volume expansion due to deuteration can be compensated for by segregation of the Cr and V. Absorption of 0.5 D/M is expected to expand the film by B3% compared to the as-deposited state (9% compared to the
segregated state). This suggests that some ‘excess’ deuterium can be stored in the first cycle, which may account for the slightly higher capacity observed in Fig. 1. The expected total expansion is 21%, which is significantly lower than what was observed, which can be accounted for by an increase in the grain boundary volume as the grain size was strongly reduced upon hydrogenation.
It would be very interesting to perform NR during multiple cycles. As the segregation step would only occur in the first cycle, volume expansion should be instantaneous, while retaining the rapid hydrogen diffusion. As phase separation requires diffusion of metal atoms, this would also explain the observed improvement of the kinetics after the first cycle.
In summary, the effect of alloying Mg with transition metals on hydriding kinetics was studied with Neutron Reflectometry for the first time by contrasting pure Mg with an Mg–Cr–V alloy. Combined NR and XRD analysis showed that Cr and V segregate from Mg in the first hydrogenation cycle, forming a percolating network through Mg, evidenced by the strongly reduced grain size of MgH2compared to pure Mg. Formation
of a blocking MgD2 layer was prevented and deuteration
proceeded nearly homogeneously through an abundance of heterogeneous nucleation sites. Enhanced deuterium diffusion was confirmed by the rapid deuteration of the bottom Ta layer.
The research herein is made possible by a reflectometer jointly funded by Canada Foundation for Innovation (CFI), Ontario Innovation Trust (OIT), Ontario Research Fund (ORF), and the National Research Council Canada (NRC).
Notes and references
1 S. Hao and S. S. Sholl, Appl. Phys. Lett., 2008, 93, 251901. 2 P. Vermeulen, A. Ledovskikh, D. Danilov and P. H. L. Notten,
Acta Mater., 2009, 57, 4967.
3 M. Conradi, M. P. Mendenhall, T. M. Ivancic, E. A. Carl, C. D. Browning, P. H. L. Notten, W. P. Kalisvaart, P. C. M. M. Magusin, R. C. Bowman, S. J. Wang and N. L. Adolphi,
J. Alloys Compd., 2007, 446–447, 499.
4 B. Vigeholm, K. Jensen, B. Larsen and A. S. Pedersen, J. Less
Common Met., 1987, 131, 133.
5 Z. Z. Karger, J. Hu, A. Roth, D. Wang, C. Ku¨bel, W. Lohstroh and M. Fichtner, Chem. Commun., 2010, 46, 8353.
6 K.-F. Aguey-Zinsou and J.-R. Ares-Ferna´ndez, Energy Environ.
Sci., 2010, 3, 526–543.
7 P. Adelheim and P. E. de Jongh, J. Mater. Chem., 2011, 21, 2417–2427.
8 S. Hao, Appl. Phys. Lett., 2010, 97, 111905.
9 W. P. Kalisvaart, C. T. Harrower, J. Haagsma, B. Zahiri, E. Luber, C. Ophus, E. Poirier, H. Fritzsche and D. Mitlin, Int.
J. Hydrogen Energy, 2010, 35, 2091.
10 H. Wipf, Topics in Applied Physics vol. 73, Hydrogen in metals III, Springer-Verlag, Berlin, 1997, ch. 3, pp. 51–91.
11 K. Higuchi, K. Yamamoto, H. Kajiioka, K. Toiyama, M. Hona, S. Orimo and H. Fujii, J. Alloys Compd., 2002, 330–332, 526–530.
12 C. T. Harrower, E. Poirier, H. Fritzsche, W. P. Kalisvaart, S. Satija, B. Akgun and D. Mitlin, Int. J. Hydrogen Energy, 2010, 35, 10343.
13 A. Baldi, M. Gonzalez-Silveira, V. Palmisano, B. Dam and R. Griessen, Phys. Rev. Lett., 2010, 102, 226102.
14 H. Fritzsche, C. Ophus, C. T. Harrower, E. Luber and D. Mitlin,
Appl. Phys. Lett., 2009, 94, 241901.
15 P. Hjort, P. Krozer and B. Kasemo, J. Alloys Compd., 1996, 237, 74–80.
16 Available from: www.ncnr.nist.gov/resources/n-lengths. 17 L. G. Parratt, Phys. Rev., 1954, 95, 359.
Fig. 4 XRD patterns of (a) as-deposited 50 nm Mg–10%Cr–10%V, (b) as-deposited 500 nm Mg–10%Cr–10%V tilted to w = 601, (c) hydrogenated 50 nm Mg and (d) hydrogenated 50 nm Mg–10%Cr–10%V.
Downloaded by CISTI Archive Access on 23 March 2011
Published on 02 March 2011 on http://pubs.rsc.org | doi:10.1039/C0CC05501H
